It is impossible to apply separation inside 4d anyway. Since we are unable to apply speed control past 4d, there's no method of doing so other than adding half a mile and hoping for the best, which doesn't seem like much of a safety case.
The point about wake turbulence is that it dissipates over time, not distance. The distances we apply are approximations that roughly correspond to the time taken for the wake to dissipate sufficiently at normal approach speeds. Since both aircraft in a pair will slow down after passing 4d, the time gap is maintained even if the distance isn't, so it really isn't an issue.